The naturally occurring androgenic hormones, testosterone and its 5α-reduced metabolite, dihydrotestosterone, are synthesized by the Leydig cells of the testes and circulate throughout the body where they diffuse into cells and bind to the androgen receptor protein (“AR”). Androgens, acting through their receptor, stimulate development of the male genitalia and accessory sex glands in the fetus, virilization and growth in the pubertal male, and maintenance of male virility and reproductive function in the adult. The androgen receptor, together with other steroid hormone receptors constitute a family of trans-acting transcriptional regulatory proteins that control gene transcription through interactions with specific gene sequences.
When prostate cancer is found to be confined to the prostate gland, the treatment of choice is surgical removal. However, 50 to 80% of prostate cancer patients already have metastases at the time of diagnosis. Most of their tumors (70 to 80%) respond to the removal of androgen by castration or by suppression of luteinizing hormone secretion by the pituitary gland using a gonadotropin releasing hormone analogue alone or in combination with an anti-androgen. The degree and duration of response to this treatment is highly variable (10% live<6 months, 50% live<3 years, and 10% live>10 years.) Initially cancer cells regress without androgen stimulation, but ultimately the growth of androgen independent tumor cells continues (35). At present it is not possible to predict on an individual basis which patient will respond to hormonal therapy and for how long. If poorly responsive patients could be identified early, they could be treated by alternative forms of therapy (e.g. chemotherapy) at an earlier stage when they might be more likely to respond.
Studies on androgen receptors in prostate cancer have suggested that a positive correlation may exist between the presence of androgen receptors in cancer cells and their dependence on androgenic hormone stimulation for growth. (An analogous situation exists in mammary carcinoma where there is a correlation between estrogen receptors and regression of the tumor in response to estrogen withdrawal). However, methodological problems in the measurement of androgen receptors have prevented the routine use of androgen receptor assays in the diagnostic evaluation of prostate cancer. Prior to our preparation of androgen receptor antibodies, all androgen receptor assays were based on the binding of [3H]-labeled androgen. These assays have been unreliable in human prostate cancer tissue because of the extreme lability of the androgen binding site and the presence of unlabeled androgen in the tissue. Endogenous androgen occupies the binding site on the receptor and dissociates very slowly (t ½ 24–48 hr at 0C). A further problem is that biopsy samples are quite small, making it difficult to obtain sufficient tissue for [3H−]-androgen binding assays. Moreover, prostate cancer is heterogenous with respect to cell types. Thus within a single biopsy sample there is likely to be an uneven distribution of cells containing androgen receptors.
Development of the male phenotype and maturation of male reproductive function are dependent on the interaction of androgenic hormones with the androgen receptor protein and the subsequent function of the receptor as a trans-acting inducer of gene expression. It has become well established over the past twenty-five years that genetic defects of the androgen receptor result in a broad spectrum of developmental and functional abnormalities ranging from genetic males (46,XY) with female phenotype to phenotypically normal males with infertility. Isolation of the structural gene for the androgen receptor makes it possible to define the nature of these genomic defects in molecular terms. Analysis of the functional correlates of the genetic defects may lead to a better understanding of the regulation of androgen receptor gene expression and of the mechanism of androgen action in male sexual development and function.
The androgen insensitivity syndrome, known also as testicular feminization, is characterized by an inability to respond to androgen due to a defect in the androgen receptor, the protein that mediates the action of androgen within the cell. Androgen insensitivity is an inherited X-linked trait that occurs in both complete and incomplete forms. The complete form results in failure of male sex differentiation during embryogenesis and absence of virilization at puberty. The result is a 46,XY genetic male with testes and male internal ducts. The testes produce normal amounts of testosterone and Mullerian inhibiting substance. Consequently development of the uterus is inhibited as in the normal male. Because of the inability to respond to androgen, the external genitalia remain in the female phenotype with normal clitoris and labia. A small vagina develops from the urogenital sinus and ends in a blind pouch. At puberty feminization with breast development and female contours occur in response to testicular estrogen, however, there is no growth of sexual hair even though circulating testosterone concentrations are equal to or greater than levels in the normal male.
Incomplete forms of the androgen insensitivity syndrome include a spectrum of phenotypes resulting from varying degrees of incomplete androgen responsiveness. At one extreme, individuals have mild enlargement of the clitoris and sparse pubic hair. The opposite extreme is characterized by more complete masculinization with varying degrees of hypospadias deformity but predominantly a male phenotype. It has been reported that some adult men with severe oligospermia or azoospermia who are otherwise normal, have defects in the androgen receptor. These may include as many as 10% of infertile males.
The genetic defect eliciting this range of abnormalities is thought to be a single biochemical event at the level of the gene for the androgen receptor. The androgen receptor is a high affinity androgen binding protein that mediates the effects of testosterone and dihydrotestosterone by functioning as a trans-acting inducer of gene expression. For proper male sexual development to occur, there is a requirement for androgen and its receptor at a critical time during embryogenesis and during puberty. The majority of individuals with the androgen insensitivity syndrome have a history of affected family members, although about a third are thought to represent new mutations of this X-linked disorder. The incidence ranges from 1 in 20,000 to 60,000 male births.
In studies of families with clinical evidence of the androgen insensitivity syndrome, four main categories were recognized that range from the most severe, complete absence of receptor binding activity in a genetic male with female phenotype, to qualitatively normal receptor in affected individuals. Second in severity are affected individuals with qualitatively abnormal androgen binding by receptor present in normal levels. Examples include the failure of sodium molybdate (a reagent often used in studies on steroid receptors) to stabilize the receptor of affected individuals when molybdate is known to stabilize the wild-type receptor. Lability of the receptor under conditions that normally cause transformation has also been reported. A third group expresses a decreased amount of receptor with wild-type in vitro binding characteristics. The final grouping contains those androgen insensitivity patients in whom no abnormality in receptor is detected. In a recent study of this form of the syndrome, the androgen receptor was as capable of binding oligonucleotides as the wild-type receptor. Indeed, with the techniques available until only recently, it has been difficult in certain cases to document an androgen receptor defect in affected individuals.
Experimental methods used in assessing receptor defects in the past have relied on the ability of receptor to bind androgen with high affinity. The limitation of this methodology is that it is not possible to distinguish between the lack of expression of the receptor and loss of androgen binding activity. An example of how inadequate methodology complicates diagnosis is the absence of detectable receptor binding activity in patients who are partially virilized. It is theoretically possible for a mutation to occur which allows the receptor with defective androgen binding activity to induce gene transcription. Biologically active truncated forms of the glucocorticoid receptor that lack steroid binding activity but retain the DNA binding domain have been demonstrated using genetically engineered mutants.
Purification of the androgen receptor has been difficult to accomplish due to its low concentration and high degree of instability. Reported attempts at purification using either conventional methods of column chromatography or steroid-affinity chromatography have yielded insufficient amounts of receptor protein to allow even the preparation of monoclonal antibodies.
An early report on the partial purification of the androgen receptor was disclosed by Mainwaring et al. in “The use of DNA-cellulose chromatography and isoelectric focusing for the characterization and partial purification of steroid-receptor complexes,” Biochem J, 134, 113–127 (1973). They used DNA-cellulose chromatography and isoelectric focusing to isolate the receptor from rat ventral prostate and determined its physiochemical properties. This group was among the first to attempt the use of steroid affinity chromatography in conjunction with conventional chromatography, using the affinity label 17B-bromoacetoxytestosterone in receptor purification (See Mainwaring et al., “Use of the affinity label 17B-bromoacetoxytestosterone in the purification of androgen receptor proteins,” Perspectives in Steroid Receptor Research, (1980)). Partial purification of androgen receptor has also been attempted from other tissue sources, such as ram seminal vesicles (See Foekens et al., Molecular Cellular Endocr, 23, 173–186 (1981) and Foekens et al., “Purification of the androgen receptor of sheep seminal vesicles,” Biochem Biophys Res Comm, 104, 1279–1286 (1982)). The partially purified receptor displayed characteristics of a proteolyzed receptor, but a purification of 2,000 fold was reported with a recovery of 33% (See Foekens et al., “Purification of the androgen receptor of sheep seminal vesicles,” Biochem Biophys Res Comm, 104, 1279–1286 (1982)). Later attempts at purification continued to combine steroid affinity chromatography with conventional techniques, reportedly achieving significant purification, but recoveries too low for further analysis (See Chany et al., “Purification and characterization of androgen receptor from steer semenal vesicle,” Biochemistry 21, 4102–4109 (1982), Chany et al., “Purification and characterization of the androgen receptor from rat ventral prostate,” Biochemistry 22, 6170–6175 (1983) and Chang et al., “Affinity labeling of the androgen receptor in rat prostate cytosol with 17B-[(bromoacetyl)oxy]-5-alpha-androstan-3-one,” Biochemistry 23, 2527–2533 (1984)). More recent studies examine the effectiveness of a variety of immobilized androgens for their ability to bind the androgen receptor (See De Larminat et al., “Synthesis and evaluation of immobilized androgens for affinity chromatography in the purification of nuclear androgen receptor,” The Prostate 5, 123–140 (1984) and Bruchovsky et al, “Chemical demonstration of nuclear androgen receptor following affinity chromatography with immobilized ligands,” The Prostate 10, 207–222 (1987)). Despite these efforts, the receptor has not been purified to homogeneity and the quantities of purified androgen receptor obtained have been insufficient for the production of antisera.
Clinical assays for the androgen receptor now include several methods. The most common is the binding of tritium-labeled hormone and measurement of binding using a charcoal adsorption assay. Either a natural androgen, such as dihydrotestosterone, or synthetic androgen, such as mibolerone or methyltrienolone (R1881), can be used. An advantage of the latter in human tissue is that it is not significantly metabolized and does not bind to the serum androgen binding protein, sex steroid binding globulin. A limitation of radioisotope labeling of receptor is interference caused by endogenous androgen. Although exchange assays for the androgen receptor have been described (See Carroll et al., J Steroid Biochem 21, 353–359 (1984) and Traish et al., J Steroid Biochem 23, 405–413 (1985)), their effectiveness is limited by the slow kinetics of dissociation of the endogenous receptor-bound androgen.
Another method used to assess receptor status is autoradiography. In this method disclosed in Barrack et al., “Current concepts and approaches to the study of prostate cancer,” Progress in Clinical and Biological Research, 239, 155–187 (1987) the radioactively labeled androgen is incubated with slide-mounted tissue sections of small tissue biopsy specimens which are then frozen, sectioned and fixed. Nuclear localization of radioactivity is detected by exposure of tissue sections to x-ray film. This technique requires considerable technical expertise, is labor intensive, and requires extended periods of exposure time. It is therefore of limited usefulness in general clinical assays. Another problem is high levels of background signal, i.e. a high noise/signal ratio, making it difficult to distinguish receptor-bound nuclear radioactivity from unbound radioactivity distributed throughout the cells.
WO 87/05049(Shine) discloses a method for the production of purified steroid receptor proteins, specifically estrogen receptor proteins, through the expression of recombinant DNA encoding for such proteins in eukaryotic host cells. However, the reference does not disclose the sequence for androgen receptor protein, nor does it disclose a method for obtaining such a sequence.